AI is quickly changing and shaping how businesses engage online. As it transforms the way in which content is created, moderated, and consumed on social media, many companies find themselves challenged to keep up. On the surface AI appears to bring incredible benefits to companies, but there are many concerns that must be addressed in order to ensure that AI is utilized effectively and safely. Here we dig into a few challenges that have come out of the rise of AI.
- Can AI create valuable content for businesses? AI tools such as DALL.E, Midjourney, Lena’s, and ChatGPT can help users create visually unique images and text. Even video content can be generated with ease. This allows creators and marketers to produce high quality media quicker than ever before.
But is the content produced through AI what viewers find value in? Critics are concerned that the content produced through AI is not authentic and can devalue a brand in the eyes of consumers. According to The Guardian, “Many of these videos qualify as ‘AI slop’ which refers to low-quality, mass-produced content that is surreal, uncanny, or simply grotesque.” Specifically, YouTube has had some issues with the amount of ‘AI slop’ that is being produced and posted on their platform.
- Are AI algorithms reliable and trustworthy in terms of moderating content on social media? Not only can the content created be quickly customized through AI, but AI algorithms can power highly tailed content on platforms like TikTok, Instagram, Facebook, and Twitter. AI tailors what users see based on their liking patterns and watch times. This targeted ad placement allows companies to optimize reach and ROI for their brands.
On the flipside, platforms are relying on AI to moderate harmful content, but sometimes these systems can misfire. It can also filter posts incorrectly and hide videos that need to be seen as well as misclassify and show posts that might be sensitive to certain viewers. In some cases that are considered high-profile, AI chat bots have produced shocking things that violate company policy, spread misinformation, or do not accurately account for minors who might be viewing the content.
2. What will the effect of AI generated profiles and virtual people be on social media? Emerging now are AI generated profiles and virtual people. Meta is expanding into deploying AI user profiles with complete bios and photostat that make content and can interact like real users. There are also experimental platforms, such as Series, which uses AI driven chat bots as social connections.
But will these fake accounts lead people to leave these networks due to a lack of authenticity and real connection? How will the content shared by these accounts be overseen to ensure that it is appropriate for the online platform, users interacting with it, and brand voice overall?
3. How will companies protect their employees and their brand from deepfakes and misinformation? A study by Business Insider used chat bots in a simulated social network. The study shows that even without promoting an algorithm, echo chambers and extreme amplification still happens. In other words, fake information can quickly go viral and be difficult to contain.
To combat this, safety features and tools are emerging. YouTube introduced an AI system that estimates user ages based on behavior to help with safety and content restrictions. Default protections still apply if they seem to be under the age of 18. But can these safety features stay ahead of the problems? Who is ultimately accountable for any misinformation spread by AI? Legally, how will laws be amended to account for the implications of AI?
As we as a culture wrestle with the ethical and moral implications of AI, the concerns raised by users have also led to the development of brand new platforms. Platforms like Cara are gaining popularity by offering users features that protect human-made content from AI interference.
AI is not just influencing social media. It is reshaping platforms and the way businesses will use them to engage with their community moving forward.
